Clinical and biochemical outcomes following structured patient education in hemodialysis: a Kenyan secondary analysis

Background Patients on maintenance hemodialysis must manage fluid, diet, and medication largely on their own to limit interdialytic weight gain (IDWG) and biochemical abnormalities. A structured educational program previously improved adherence in this cohort, but its association with objective outcomes had not been reported. Methods This secondary analysis of a quasi-experimental study (three Kenyan dialysis centers in Thika, Nairobi, and Kisumu, March–June 2025; one intervention facility, two pooled control facilities) included 115 of 129 participants (46 intervention, 69 control). IDWG, expressed as a percentage of dry weight, and serum sodium, potassium, phosphorus, urea, and creatinine were compared using analysis of covariance (ANCOVA) adjusted for baseline value, age, sex, and education. Within-group changes were examined with paired t-tests, and baseline characteristics were compared with Welch’s t-tests and Fisher’s exact test. Results Groups differed at baseline in serum sodium and potassium (both p < 0.0001). In the adjusted models, the intervention group had lower endline IDWG (difference −1.18 percentage points of dry weight, 95% CI −1.86 to −0.50, p = 0.001) and lower endline sodium (difference −4.69 mmol/L, 95% CI −6.57 to −2.81, p < 0.0001) than control. Serum potassium was significantly higher, not lower, in the intervention group (difference +0.49 mmol/L, 95% CI 0.27 to 0.71, p < 0.001); the difference remained positive across the observed baseline range (+0.40 to +0.66 mmol/L), although a significant baseline-by-group interaction showed that its size varied with baseline value. Phosphorus, urea, and creatinine showed no overall group association; urea and creatinine also had significant baseline-by-group interactions, so their association with group varied with baseline value. Conclusion Structured education was associated with lower IDWG and lower sodium, but with higher, not lower, adjusted serum potassium; phosphorus, urea, and creatinine showed no consistent overall association. The lower sodium should not be read as a benefit, because the intervention group’s mean endline concentration was below the conventional reference range, and the potassium finding does not support a beneficial effect on potassium control. Because intervention status was fully determined by facility, these findings are adjusted associations, not causal intervention effects, and need confirmation in multi-site or randomized designs.